This week, top on everyone’s mind has been the banking sector. Fourth quarter results indicate a stomach churning rise in the number of bad loans or loans which borrowers are not paying back. At such a time, the regulator Reserve Bank of India is imposing sterner rules called the Basel rules.
These rules require banks to back up their deposits and loans with more and more shareholder capital. How will banks tackle this double whammy and is the RBI right in demanding higher safeguards whatever the economy's condition? Also, how bad is the bad loan problem?
RBI Deputy Governor Anand Sinha, State Bank of India Managing Director Diwakar Gupta and ICICI Bank Executive Director NS Kannan join CNBC-TV18’s Latha Venkatesh to discuss how banks will tackle with this double whammy of Basel III norms and rising NPAs.
Below is an edited transcript of their interview. Also watch the accompanying video.
Q: The bad loan numbers plus the restructured assets are beginning to look a little scary. How does the RBI look at this issue and how should investors understand this problem of restructured assets, because now it is beginning to look like eroding the net worth?
Sinha: No, it is not correct to say that so much of assets are taking away the net worth. As far as the NPAs are concerned, there is a provisioning rule and it’s not 100% unless it’s a loss asset. Therefore, it’s not really taking away the entire net worth.
To my mind, restructuring is a very legitimate and a desirable tool to deal with a stress situation. In certain parts of the world, more particularly Asia and that is where we were also prior to 2001, any restructured asset would be treated as a NPA and then after a defined period of performance according to the rescheduled installments, you could upgrade it.
Then after that, we switched over to another system, which was more prevalent in the west, where if you are fairly confident that it will be able to meet its rescheduled liabilities, then you keep it as standard. What you do is you fair value it and the loss in the net present value you provide for in the balance sheet. So this is the system that we are following now.
The question is whether restructured assets are really NPA, and I mean this from an economic sense not a classification sense. The evidence that we have so far is that roughly 15% restructured assets do turn into NPA. In some banks it is much lower, that’s what we understand, but overall 15% does turn into NPA. 85% is not NPA, therefore it should not be a threat to the economy.
You have two ways now to deal with that. Either you downgrade all the restructured assets into NPA and then take up 85-80% after the defined period into the standard asset category, or you follow the other method.
Q: Has RBI defined and allowed banks to take out their restructured assets from that category? The debate was on, but I understand you’ll have not yet allowed them to pull those assets out of restructured category?
Sinha: I think there is a bit of semantics involved over here, so let me try and clarify. As per the present dispensation, once an asset is restructured, it is to be shown as a restructured asset maybe for all times as of now. But from asset classification perspective, this categorisation doesn’t carry because then it becomes a standard asset once it has passed the probation period.
Banks made a representation to us that once a restructured standard asset has passed the test, then why continue to call it a restructured standard asset and we have agreed for that.
Q: The numbers that have come for FY12 in the form of NPLs and restructured assets for many of the banks has been quite stomach churning. Is this in your estimate the last of the bad quarters or are we going to see couple of more bad quarters?
Gupta: I think there is no denying that industry and business in general has been through trying times for two years and more. Obviously staying capacity of people varies; some people can ride it out, some people try and some people are not able to.
I think the stress that we are seeing is really systemic and it’s more because of external factors. If I were to go by the anecdotal data that we are observing, we have actually passed the worst. Q2 and Q3, were very bad and going forward we should see a significant improvement in the scenario.
Q: Would you share Mr Gupta’s optimism, and not just for ICICI since you have over 10% of banking industries assets?
Kannan: I would broadly agree with that assessment. If you look at our own numbers to start with, in the month of December and January when we came out with the results for the third quarter, I had mentioned in that point in time itself that restructuring will continue and you will see that number going up in Q4. It has pretty much panned out the way I had forecasted at that point in time.
If we look at the indicators at that point in time, we saw more references to the corporate debt restructuring mechanism. We had seen the access to capital market of some of the companies got impaired because of the equity market situation, so this was anyway coming.
But when we look at the numbers now, having put out the March results, I would agree with Mr Gupta that the bulk of restructuring is behind us. I don’t have a pipeline like I had in December to make that statement, so I would say that broadly the issue is behind us.
Q: If I ask you to extrapolate for the industry?
Kannan: My sense is since October-November, I haven’t seen large scale fresh reference to corporate debt restructuring schemes (CDR). So I would say broadly that would be true for industry as well, but we will have to wait and see how it pans out.
Only one point I want to supplement what Mr Gupta mentioned. The NPLs and restructured assets are qualitatively different, that is a point I would like to make, and that is born out of our own experience as well.
If you look at the restructured assets, it will work only when fundamentally the business model is viable and you have some short-term problems such as the cash flow issues. But if you are fundamentally going to restructure a company which is not viable, it is not going to work. It is going to get shown up as an NPA within few quarters. So it’s only a postponement of problem.
If you look at our own restructures assets, the slippage into NPL is less than 5%. So I would suggest that these two categories should qualitatively look different.
Q: A bunch of banks still indicate higher aggregate NPLs and NPLs as a percentage of total assets. So does the regulator feel confident that the worst is behind us?
Sinha: Regulators surmise is dependent on what we discuss with the banks and what independent input we have. So maybe a month or two back, we had a very elaborate exercise in which we had looked at the NPAs of the system, we had extensive talks with the banks. The sense we got was that NPAs would increase in the near-term, but after some time they would not really be on any increasing spree.
Now you cannot take out any conclusion out of the context if the macro economic situation as envisaged in the monetary policy were to change substantially. But I would believe based on that discussion, and I can also tell you we are going to have another discussion shortly, it seems that the NPAs should not be rising in a significant amount or by any significant percentage.
Q: Is there any part of the Basel III norms that you would describe as perhaps particularly unkind or that might require a change? The fact that Tier-II capital is largely long-term debt which should be loss bearing has kind of frightened banks and some of them said that hereafter we will raise only common share as capital, we will not even raise Tier-II. Would that be one of the bigger problems?
Gupta: I think the leverage ratio in Basel III is likely to be a bigger challenge to banks than the capital ratios per se, because they limit the balance sheet. For the leverage ratio there is no risk waiting. I think the intention of the leverage ratio also is to see that our balance sheet exposures remain controlled but for now if State Bank were to implement Basel today, we would need capital for leverage rather than for capital adequacy on the risk weighted asset.
Q: How much would you need? What is the leverage?
Gupta: It’s very marginal. We need about 4.5% leverage. We are somewhere in the high 3.5-4%. We may need about Rs 2,000 crore today. But this amount could go up to about Rs 45,000 crore by 2018 if we assume a 20% growth rate.
Q: Could getting capital from the shareholder therefore become a little more onerous?
Gupta: Yes. The government has to provide the capital and therefore it has to be provided for in their finances. Otherwise, I don’t think it’s really going to be that big a problem to raise a lakh of crore in six or seven years.
Q: We all know that the government has other onerous commitments and therefore finding capital even for one of the banks is a very big problem and a whole host of them. Would you care to tweak any of these rules?
Sinha: As far as pension is concerned, the whole thing in the last settlement was not managed properly. As a result, we had to tweak the rules which in fact we did. Therefore, once we shift to Basel III now we have to wipe the slate clean. I don’t think there will be further tweaking possible.
Q: The government is in sync with this requirement?
Sinha: This is a prudential requirement. We haven’t talked to the government on this. This is purely a prudential requirement.
Q: It is globally demanded?
Q: What would be an issue for private sector banks? Obviously pension is not an issue and raising shareholder capital is not as serious a problem for private sector. Would it be this equity investment in subsidiaries?
Kannan: If you look at a financial conglomerate like ICICI Bank, here we do have our subsidiaries, ICICI Bank is also the holding company of the ICICI Group, and we do have investments in our subsidiaries. Under Basel II hitherto we were doing the subsidiary investments which is deducted equally from Tier-I and Tier-II.
That’s the way we compute our Tier-I and Tier-II capital adequacy ratio. With the application of Basel III they have what is called a corresponding direction approach. That means that if the subsidiary investment is treated as equity by the subsidiary then it should be removed from Tier-I directly. So to that extent there will be a bit of an impact.
It’s not that everything is applicable from 2013 or 2014 and you have this spread out till 2018. So if you put it in that context, we have a Tier-I capital adequacy of 12.68% plus also the fact that the threshold levels are much higher in Basel III, even then the next three years we don’t have to access capital for growth. So we don’t have any issues really with the application of Basel III.
Q: But you will also have the capital conservation buffer kicking in, maybe counter cyclical at some point?
Kannan: Yes, the capital conservation buffer we are talking about 2.5% now and that also again has a face in arrangement. It’s not that 2.5% comes in. So it gives an enough timeframe for banks like us to look at our business models and say - I have also looked at prima facie numbers on the leverage ratio. We don’t have any issues. We are well above 4.5% we don’t have any issues.
Q: So would you have an issue at all with respect to some rules that could be perhaps more benign?
Kannan: On Basel III itself if you really look at the world over how things have moved. The focus is so much on increasing capital requirements, reducing leverage and more importantly, the quality of capital. I don’t think you can sort of fight these trends. That is something which is important from a prudential perspective. I guess all of us will have to look at the business models to align itself to that. We will have to wait and see how the Tier-II market itself develops because the basic requirements which have been put in Tier-II is of a minimum maturity of 10 years and it has to have to some loss absorbing capacity.
So those kind of instruments are really new. We will have to wait and see how that market really develops. But my own interaction with investors and analysts over the last two-three years is where people largely focus on Tier-I. They don’t really look at the total capital. In assessing the bank’s quality, they look at Tier-I more closely.
Q: But don’t you think your growth will be constrained if you cannot raise Tier-II capital?
Kannan: I guess each of the banks will be focusing so much on Tier-I and the ability to raise common equity which is what is going to determine growth rather than Tier-II.
Sinha: On Tier-II let me clarify something. Perhaps you said that Tier-II doesn’t have loss absorbing capacity on something?
Q: No, since it is forced on it, will there be a market?
Sinha: No, let me clarify. The capital structure has two parts. The first part that is Tier-I has to absorb losses on an ongoing basis. Tier-II is very much a loss absorbing capital, but only in the case of liquidation. So what happens is that under Basel III there has to be a clause that if the bank reaches a situating where it’s likely to go down under, then at that stage Tier-II has to be converted into equity to be able to absorb the losses.
Similar condition with Tier-I but that is more in an ongoing kind of situation. There is a trigger point for a bank which is running. So ultimately both of them have to convert to equity but with different intent. Tier-II is very much a loss absorbing capital. The only thing that it does not support is absorption of loss in a non-liquidation situation.
Q: Would there not be an impact on the RoE because a lot of shareholder capital will have to be raised. So RoEs are under pressure don’t you think?
Gupta: Certainly it will affect the RoE considering that you will need more own funds, more equity and return is going to fall. But these are all international benchmarks, there is nothing sacrosanct about a RoE number. Today 18% is fine, 14% is low, and internationally 8 or 9% of RoE is good and now again that is benchmarked with a return that you would get on a deposit. So there are several things, if we need more capital as a prescription and people will be happy with lower RoE. I think industry will accept it.
Q: Would you worry on RoE?
Kannan: I would say that systemically if we look at the kind of capital threshold increases which have been put in place, clearly it will have probably a 1-2% kind of impact on RoE that cannot be wished away. But on other side, one could argue saying that you are reducing risk, so if you look at the risk return equation maybe we are neutral, so that is the only point.
Q: What you are saying in dynamic provisioning is that under stressed times, in the last 10 years you have found that banks lose 1.4% of their capital and therefore that has to be perennially provided for by banks till they reach 1.4% - how much more expensive would that be, is that too tall an order?
Gupta: I think the discussion paper put out by RBI is really a discussion paper. We don’t need to worry too much about the numbers. For example, probability of default in their initial paper are calculated on the stock rather than the number of accounts. Similarly, the loss given default is on the basis of just write-offs and not on the basis of recoveries in written-off accounts for example.
So there is a lot on data which can change the numbers that have been derived but the 2-3 basic issues on dynamic provisioning that we need to address ourselves to is that we should be providing for downturn loss given default (LGD) or for normal LGD because expected losses are any way across a cycle.
It is really a theoretical debate on what we should be looking at rather than the number because the number I am sure will change from bank to bank and even for the pool. Even today for the nine banks the data which the RBI has relied on comes out of Off-site Monitoring & Surveillance System (OSMOS) and I know the data integrity at least for some of the banks and mine is included in them, is not up to the mark. Therefore, the results will change.
Q: What would be your biggest problem with dynamic provisioning, you have to provide such a lot of money and they will tax you also on it?
Kannan: I would like to talk about two-three broad issues with the discussion paper. It is in form of a discussion paper and like what the RBI has done, they have a committee then they put out a discussion paper, and we are given an opportunity to go back to them and in this case we have an opportunity up to May 14, so then we will do that. If we look at a couple of issues there, firstly the write-off has been used as a proxy for computing the LGD. So here, write-off itself has got several dimensions to it, there is a tax dimension, there is internal policy of each of the banks and again sometimes recovery back from the write-offs.
Maybe from the RBI’s perspective, the data availability may have resulted in write-offs being looked as a proxy. That is something which we would believe that a better estimate from the banks themselves on what has been the LGD in case of defaulted account is something which is better estimated to go by. If you looked at the four categories which have been put out in the discussion paper, you have mortgages, corporate loans other than infrastructure, then other retail, then you have ‘others’ as a category which is some sort of balance category.
There you have things starting from commercial real estate, to staff loan, to SME loan, to infrastructure loans. So all of them have been clubbed under a category. Again, there may have been data available issues which may have resulted in that kind of categorisation. If you look at the kind of provisioning requirement we are talking about something like 2.3% per annum for that categorisation.
Now if we look at our own actual data, the infrastructure has not led to that kind of default situation. To supplement Mr Gupta’s point, if we look at the LGD for an infrastructure as a proxy for that category, we are looking at a 90% as a loss given default should a default happen. So this is one area where we can look at. A lot of these problems will get resolved when you do a dynamic provisioning based on internal data of the banks.
But now one has to see what precedes what because the internal data will get more and more robust as we move to the advanced approaches under Basel II. As and when that happens, then the internal data for both will be available. If we apply that data then there shouldn’t be any divergence between the two methods. So one has to wait and see which comes first and which comes later.
Q: I don’t think you are going to allow banks to have internal rating based provisioning up until maybe 2014 or thereabouts, it’s a long process?
Sinha: No, nothing like that. It has taken us almost two years to put out this paper. So now we have given it back to them that if you are frightened with these numbers then come out with better numbers and better methodology. You also see that the entire paper is based on data from only nine banks.
Q: The next step is you are awaiting data from the banks, so when do you think the entire process might end and you may say that dynamic provisioning starts from this date?
Sinha: I won’t like to put any timeline on this because it needs to be made much better and that is why in the discussion paper we have raised many questions, we have asked the banks to give their own estimates. So ultimately once the banks find that the figures we have come out with is uncomfortable, I am sure that it will provide enough incentive for them to look into the data, come back to us with better data and maybe suggest better methodology also.
Q: From an investor or a banker point of view at least FY13 will not be the moment when all this will have to be provided for?
Sinha: No, this is only a discussion paper so it has to undergo many stages now. First the banks have to come back, then we have to study all that they say. It’s a very consultative process.
Q: So we can worry about it in FY14 or rather end of FY13.
Sinha: Certainly not in the immediate future, may not be for six months at least.